摘要
一般常用的边缘检测算法计算量较大,对细小边缘检测效果较差,不适合实时边缘检测系统。针对这一问题,本文提出一种连续分割的快速边缘检测算法:从相互垂直两方向分割梯度图像,提取截面曲线极大值获得图像边缘,使用形态学梯度,检测出细小的图像边缘。实验结果表明此算法较Canny等经典算法减少了计算量,提高了边缘检测精度。
The general edge detectors, such as Canny, snake etc., are computationally expensive and can＇t detect the imperceptibility edges well. A fast edge detection technique using successive segmentation is presented, The imperceptibility edges are detected well by using the morphological gradient. The experimental results show that our proposed technique performs favorably when compared with many other well - known edge detection algorithms. This edge detection technology can be used widely in system such as ART, real time object tracking, etc,
出处
《电讯技术》
2006年第1期88-93,共6页
Telecommunication Engineering
关键词
视频图像
边缘检测
快速算法
连续分割
形态学梯度
video image
edge detection
fast algorithm
successive segmentation
morphological gradient
